201005734 九、發明說明： 【發明所屬之技術領域】 本發明係關於冗餘資料儲存方法及系統。更特定言 本發明係關於冗餘電子資料儲存方法器件及系統^ 中斷回復。 罵入 【先前技術】 • -類型之電子資料储存系統使用以獨立磁碟冗餘 (獅)格式所配置之各式各樣磁碟機，資料係橫跨複固 ❿ 磁碟鏡像。在此類配置中，若-資料儲存磁碟變得不可 用，則可以從其他磁碟之一存取資料。此類資料儲存系統 常常係稱為η方向鏡像系統。 在許多η方向鏡像系統中，（例如）由系統斷電引起之寫 入中斷可以使資料儲存系統處於一其中僅將新資料寫入至 資料儲存器件之子集之狀態下。若未偵測及校正此類狀 況，則危及η方向鏡像系統之完整性，因為不再保證每— 資料儲存器件包含其中所儲存之相同資料》 存在用於在資料儲存器件之η方向鏡像集中偵測及從寫 入中斷回復之習知方法。不過，一旦已偵測到寫入中斷校 正寫入中斷時習知方法可能相對低效。例如，通常，不儲 存用以指示寫入操作期間各儲存器件上之什麼區正在改變 之資訊。為了從中斷回復’習知方法常常實行來自一器件 之所有資料完整複製至所有其他器件。或者，習知方法實 行所有器件上之所有資料的無一遺漏比較以決定差異，接 著义/員校正該等差異。兩個程序係相對低效的。此外，隨 135210.doc 201005734 著器件之大小或器件之教 線性增加。 目增加 此等習知程序之低效性 因此，需要用於在冗餘 統)中校正偵測到之窝由’儲存系統(例如n方向鏡像系 【發明内容】…斷的已改善方法。 涉及儲存= = 資料儲存方法、裝置及系統中，其 邊界資：該方法^ 件之-宜w 儲存針對資料至複數個資料储存器 二二界資訊，將該資料寫入至該複數個資 =件及該資料至該複數個資料健存器件之該寫入操 =元成時移除該已記錄邊界資訊。該邊界資訊可以指示其 i寫入操作期間正在寫入特定資料集之資料儲存器件 :。若該資料至一或多個資料餘存器件之該寫入操作期間 一中斷發生，則可以使用該邊界資訊藉由識別其中該中斷 發生時正在寫入資料之特定資料儲存器件區而從該中斷回 復。因此’與習知資料儲存系統不同，僅其中該中斷發生 時正在寫入資料之該等特定區需要加以重寫。 【實施方式】 在以下說明中，相同參考數字指示相同組件以藉由圖式 之說明増強本發明之瞭解。此外，儘管下文中論述特定特 徵、組態及配置，但應瞭解僅基於解說目的而如此論述。 熟習相關技術者應認識到其他步驟、組態及配置係有用的 而不背離本發明之精神與範疇。 現在參考圖1，顯示一習知冗餘資料儲存系統1〇之方塊 135210.doc 201005734 二 =儲存系統_合至一包括適於使用資料館 以'、：〇之應用程式的主機系統。資料鍺存系統1〇可 =括-資料儲存器件控制器12，其轉合至許多資料储存 ;件資：如第一資料储存器件14、第二資料储存器件邮 第二資料儲存器件18。該等資料儲存器件 資料儲存磁碟或磁碟機，或其他適合的資㈣存器件。 :::存器件控制器】2亦耦合至主機系㈣，主機系統 〃。至產生欲儲存資料之應用程式24。資料储存器件 控制器12經由主機系統22從應用程式24接收資料且將接收 到之資料儲存至資料儲存器件14、16、18之每一者藉此 几餘。例如，若應用程式24產生三個欲儲存資料集 (例如，資料集A、資料集B、資料集〇,則資料储存器件 控制器12會將各資料集寫入至資料儲存器件14、16、18之 乂此方式，若資料儲存器件之一（例如）由於磁碟 機故障、資料毀損或其他不可用狀況而變得不可用，則資 ;、儲存系統1 〇提供資料儲存g整性及存取。在此類情況 下，可以從其他資料儲存器件之一存取資料。 過甚至在此類几餘資料儲存系統中，當將資料寫入 至資料儲存器件時，寫入操作中斷、錯誤狀況、斷電或可 存取資料儲存器件之添加或移除可能發生，因此對剛剛寫 入或目前正在寫入至資料儲存器件之資料之有效性表示懷 疑。例如，藉由至儲存器件之寫入操作期間遇到一中斷， 或多個之資料儲存器件控制器12、主機系統22及應用程 式24將要決定哪一（或哪些）資料集係有效的，即，將哪些 135210.doc 201005734 器件加以儲二所有資料儲存器件並橫跨所有資料儲存 器件加以儲存。此外，若已決定中斷已予 個資料儲存器件之資料之1袖目丨咨 冑人至一或多 12或其他適當组件神至 資料儲存器件控制器 存器件^ 橫跨受中斷影響之所有資料儲 之程序。 〇復或以其他方式還原受影響資料之完整性 如上文中所論述，習知地，至資料儲存 操作期間，皋指千曰& 買科罵入 ❹ ❿ ㈣户 不目别正在寫入或剛剛寫入資料所至之資 =儲存器件之特定區或位置的資訊被寫入或以其他方式儲 存於資料儲存器件上、資料 錯 古#料儲存控制器内#其他任何地 方。用以從資料寫入中斷回復之習知 器件控制器！2或系統1〇内之其他貧科错存 存器件之-(例如，第一資料=:件通常決定資料儲 資科儲存器件14)是否其中已儲存 :::、已寫入至所有資料儲存器件之所有資料(例如，資 有資料集B及I接著，將來自該特定資料儲存器件之所 〜、之π整複本寫入至所有其他資料儲存器件。例 如，將成功儲存於第一資料儲存器件14上之 及C之完整複本複製至其他資料儲存器件，例如，第二資 :儲存器件16與第三資料儲存器件18。如本文先前所論 述’此程序係相對低效的。 或者’資料儲存器件控制器12或資料储存系統ι〇内之其 他適當組件可以嘗試僅複製橫跨資料儲存器等 資料集或資料集之部分。不過，此程序涉及實行寫入至^ 有資料儲存器件之所有資料集之無一遺漏比較的相對費力 135210.doc 201005734 程序。此類比較完成時，接著可以將 之部分從具有資料A J貝η果次貧料集 已危…ίί 來源資料儲存器件複製至可能 過此&整性之—或多個目標詩料11件。不 過此替代程序可能正像將來自具有資 料儲存器件之所有資料隼1选制 ^鳘性之來源資 n “ 有貢枓集疋整複製至可能完全不具有資料 Ϊ的話。所有目標資料儲存器件那樣低效且耗時，若非更 料=1月之具體實施例提供一種針對用於债測及從冗餘資 予:統（例如η方向鏡像資料儲存系統）之寫入操作中斷 習知方法之改善。本發明之具體實施例涉及储存寫 、之邊界資訊，即，寫入操作之邊界。例如，該邊界 訊可以指示藉由該目前資料寫入操作欲寫入 =件之哪―區。在寫入操作中斷之事件中，可以使用邊 界資訊來從中斷寫人操作回復。該邊界f訊可以儲存 料儲存器件之一 3¾¾加、 之戍多個内之一位置上或寫入至資料儲存器 或多個内之一位置及/或寫入至資料儲存器件外部 之一位置，例如資料儲存裝置及/或其控制器内。 在使用邊界資訊時，從中斷寫入操作之回復程序（例如） 藉由減少校正寫入中斷所需要之時間而比習知方法中更高 效。藉由s己錄或儲存各寫入操作之邊界，僅資料儲存器件 之寫入中斷期間正在寫入所至之區(即臨界區)需要考量 回復或校正。資料儲存器件之在臨界區外部或與臨界區不 同的區域無需改變或重寫，且因此在每一資料儲存器件上 保持相同。因此’回復時間無需花費在複製資料儲存器件 135210.doc 201005734 ^臨界區或决定複數件中 广回復程序期間僅需要複製臨界 ：：不同的 儲存系統之從中斷寫入操作 匕方式，資料 言比習知回復程序有明顯改善。相復時間而 資圖2 ’顯示依據本發明之具體實施例之-冗餘201005734 IX. Description of the invention: [Technical field to which the invention pertains] The present invention relates to a redundant data storage method and system. More specifically, the present invention relates to redundant electronic data storage method devices and systems. Intrusion [Prior Art] • - The type of electronic data storage system uses a variety of disk drives configured in the independent disk redundancy (Lion) format, and the data spans the complex ❿ disk image. In this type of configuration, if the data storage disk becomes unusable, the data can be accessed from one of the other disks. Such data storage systems are often referred to as n-direction mirroring systems. In many n-direction mirroring systems, for example, a write interrupt caused by a system power down can cause the data storage system to be in a state in which only new data is written to a subset of the data storage device. If such conditions are not detected and corrected, the integrity of the η-direction mirroring system is compromised because there is no longer a guarantee that each data storage device contains the same data stored therein. A known method of measuring and replies from a write interrupt. However, conventional methods may be relatively inefficient once a write interrupt correction write interrupt has been detected. For example, typically, information indicating what areas on each storage device are changing during a write operation is not stored. In order to respond from interrupts, conventional methods often perform a complete copy of all data from one device to all other devices. Alternatively, the conventional method performs a non-missive comparison of all the data on all devices to determine the difference, and then the sense/member corrects the difference. Both programs are relatively inefficient. In addition, the size of the device or the linearity of the device increases linearly with 135210.doc 201005734. In order to increase the inefficiency of such conventional procedures, there is a need for an improved method for correcting detected nests in a redundant system by a 'storage system (eg, n-direction mirroring system). Storage = = data storage method, device and system, its boundary assets: the method of the method - should be stored for the data to a plurality of data storage 22 information, the data is written to the plurality of assets = And deleting the recorded boundary information by the data to the plurality of data storage devices. The boundary information may indicate a data storage device that is writing to a specific data set during the i write operation: If an interruption occurs during the writing operation of the data to the one or more data remaining devices, the boundary information can be used to identify the specific data storage device region in which the data is being written when the interrupt occurs. The reply is interrupted. Therefore, unlike the conventional data storage system, only the specific areas in which the data is being written when the interrupt occurs need to be rewritten. [Embodiment] In the following description, The reference numerals indicate the same components to comprehend the understanding of the present invention by the description of the drawings. Further, although specific features, configurations, and configurations are discussed below, it should be understood that this is only discussed for illustrative purposes. Those skilled in the art will recognize Other steps, configurations, and configurations are useful without departing from the spirit and scope of the present invention. Referring now to Figure 1, there is shown a conventional redundant data storage system 1 block 135210.doc 201005734 2=storage system_to one It includes a host system suitable for use in the library with ',: 〇 application. The data storage system 1 can include - data storage device controller 12, which is transferred to many data storage; The device 14 and the second data storage device are the second data storage device 18. The data storage device data storage disk or disk drive, or other suitable resources (4) memory device. ::: memory device controller 2 is also coupled To the host system (4), the host system 至. To generate the application 24 to store the data. The data storage device controller 12 receives the data from the application 24 via the host system 22 and The received data is stored to each of the data storage devices 14, 16, 18 by a number of times. For example, if the application 24 generates three data sets to be stored (eg, data set A, data set B, data set) The data storage device controller 12 writes each data set to the data storage device 14, 16, 18 in this manner, if one of the data storage devices (for example) is due to a disk drive failure, data corruption, or other unavailability If the situation becomes unavailable, then the storage system 1 provides data storage g integrity and access. In such cases, the data can be accessed from one of the other data storage devices. In the data storage system, when a data is written to the data storage device, a write operation interrupt, an error condition, a power failure, or an addition or removal of an accessible data storage device may occur, and thus may be written or currently being written. The validity of the information entered into the data storage device is questionable. For example, by encountering an interrupt during a write operation to a storage device, or a plurality of data storage device controllers 12, host systems 22, and applications 24 will determine which (or which) data sets are valid, ie, Which 135210.doc 201005734 devices are stored in all data storage devices and stored across all data storage devices. In addition, if it has been decided to interrupt the data that has been given to a data storage device, one to more than one or more 12 or other appropriate components to the data storage device controller device ^ across all data storage affected by the interruption The program. Recovering or otherwise restoring the integrity of the affected data is as discussed above. Conventionally, during the data storage operation, 皋指曰& 骂科骂入❹ 四 (4) households do not witness the writing or just The information to be written to the data = the specific area or location of the storage device is written or otherwise stored on the data storage device, the data is wrong #material storage controller # anywhere else. A well-known device controller for writing interrupts from data! 2 or other poorly-deposited devices in the system 1 (for example, the first data =: the piece usually determines the data storage department storage device 14) whether it has been stored:::, has been written to all data storage All information about the device (for example, the resource set B and I then writes the π-replica from the specific data storage device to all other data storage devices. For example, it will be successfully stored in the first data storage. A complete copy of device 14 and C is copied to other data storage devices, for example, second resource: storage device 16 and third data storage device 18. As previously discussed herein, 'this program is relatively inefficient.' Other suitable components within the storage device controller 12 or data storage system may attempt to copy only portions of the data set or data set that span the data store. However, this procedure involves the execution of all of the data storage devices. The data set has no relative omissions compared to the relatively laborious 135210.doc 201005734 procedure. When such a comparison is completed, it can be partially taken from the data AJ shell η fruit poor material set has been dangerous... ί The source data storage device is copied to 11 pieces of the target material or multiple target poems. However, this alternative program may be like the source of all the data from the data storage device.资 n “There is a collection of tribute to the possibility that there is no data at all. All target data storage devices are as inefficient and time consuming, and if not more than one month, the specific embodiment provides a measure for debt testing and Redundancy: The improvement of the write operation interrupt method of the system (e.g., the η-direction mirror data storage system). The specific embodiment of the present invention relates to storing the boundary information of the write, that is, the boundary of the write operation. The boundary message may indicate which area of the device is to be written by the current data write operation. In the event of a write operation interruption, the boundary information may be used to reply from the interrupt write operation. The boundary information may be stored. One of the material storage devices 33⁄4⁄4, or at one of a plurality of locations, or written to a data storage device or a plurality of locations and/or written to a location outside the data storage device, for example For example, in the data storage device and/or its controller, when using boundary information, the reply program from the interrupt write operation (for example) is more efficient than the conventional method by reducing the time required to correct the write interrupt. The boundary of each write operation is recorded or stored by the s, and only the area (ie, the critical section) that is being written during the write interruption of the data storage device needs to be considered for recovery or correction. The data storage device is outside the critical region or The different areas of the critical section need not be changed or rewritten, and therefore remain the same on each data storage device. Therefore, the 'recovery time does not need to be spent in the copy data storage device 135210.doc 201005734 ^Critical area or during the decision process Only need to copy the critical:: the different storage system from the interrupt write operation, the information is significantly improved than the conventional reply procedure. In contrast, Figure 2' shows a redundancy in accordance with a particular embodiment of the present invention.
存系統眘从鏡像系統或其他適合的冗餘資料儲 =，存系統3。包括一資料_置32，資= 2包括一資料儲存器件控制器34 更詳細論述’資料儲存器件控制器34可以包括一 :訊之儲存的邊界資訊位置^此外，代 = 内之邊界資訊位㈣之外， 資料儲存， 円）之邊界資訊位置38。 資料儲存裝置32經由資料健存器件控制器34可以輕合至 複數個資料儲存器#，例如第-資料儲存器件42、第二資 §1存n件44及第二資料儲存||件46。該等資料儲存器件 之每者可以為資料儲存磁碟或磁碟機，或任何其他適合 的資料儲存ϋ件。資料儲存裝置32係經組態用以麵合至一 主機系統（未顯示）’該主機系統通常亦耦合至一產生欲儲 存於資料儲存系統30内之資料的應用程式（未顯示）。 貝料儲存裝置32及/或資料儲存器件控制器34可以部分 地或儿全地包括任何適合的結構或配置（例如，一或多個 135210.doc -12- 201005734 積體電路）。此外’應瞭解資料儲存裝置32包括其他組 件，硬體與軟體（未顯示），其係用於資料儲存裝置32及/或 資料儲存器件控制器34之本文未明確說明的其他特徵與功 能之操作。資料儲存裝置32及，或資料儲存器件控制㈣ 之所有相關部分可以部分或完全採用硬體電路及/或大器 件或組件群組内之其他硬體組件之形式加以組態。或者， 資料儲存裝置32及/或資料儲存器件控制器％ 部分可以部分或完全制軟趙之形式（例如，作j = 令及/或一或多個邏輯或電腦編碼集）加以組態。在此類組 態中，邏輯或處理指令通常係儲存於記憶體元件中。記憶 艘元件通常係轉合至處理器或控制器，例如，資料儲存器 件控制器3 4。控制器從記憶趙元件存取必需的指令且執行 指令或將指令傳輸至資料儲存裝置32内之適當位置。 現在參考圖3(繼續參考圖2)，顯示依據本發明之具髖實 施例用於寫入至圊2之資料儲存系統之一方法8 〇之方塊 圖。將隨同資料儲存系統30之操作一起來說明方法8〇。作 為冗餘資料儲存系統之部分，資料儲存裝置32經由資料儲 存器件控制器34將藉此接收到的資料儲存至資料儲存器件 42、44、46之每一者。因此，若資料儲存器件之一其後 (例如）由於磁碟機故障、資料毀損或其他不可用狀況而變 得不可用，則可以從其他資料儲存器件之正確發揮功能之 一資料儲存器件存取資料。 方法80包括將邊界資訊儲存（例如）至一或多個之資料儲 存器件及/或至資料儲存裝置32之步驟82。依據本發明之 135210.doc •13- 201005734 '、體實施例’資料儲存器件之某些或全部之-部分、資料 儲存裝置32之—部分及/或任何其他適當位置係保留用於 邊界資讯之儲存。例如’針對邊界資訊之儲存保留資料儲 存器件之 ’、二或全部之上或之内之一邊界資訊位置或區。 或者，瓦Ν 一 以在資料儲存裝置32或替代地耦合至資料儲存裝 置32與資料儲存器件42、44、46之一或兩者的一外部位置 . 内保留一邊界資訊位置或區。如上文中所論述，邊界資訊 可以指不欲將特定資料集寫入至資料儲存器件之什麼區、 ❿ 冑料儲存器件之欲寫人資料所至之區之開始位置與結束位 置、貝料儲存器件之欲寫入資料所至之區之開始位置與長 度及/或關於欲寫入至資料儲存器件42、44、46之資料集 的其他適當資訊。 例如，若藉由資料儲存裝置32接收N個資料集以用於儲 存至資料儲存器件42、44、46之每一者則資料儲存器件 控制器3 4 (或（例如）資料儲存裝置3 2内之其他適當組件）保 φ 留一邊界資訊區或位置用於與N個資料集儲存至個別資料 儲存器件及/或至全部資料儲存器件相關之邊界資訊之儲 存。例如，資料儲存器件控制器34在第一資料儲存器件42 中保留一第一邊界資訊區51用於與N個資料集儲存至第一 資料儲存器件42相關之邊界資訊之儲存。或者，資料儲存 器件控制器34於其中保留邊界資訊區36及/或在資料儲存 裝置32内保留邊界資訊區38用於與n個資料集儲存至第一 資料健存器件42相關之邊界資訊之儲存。此類邊界資訊可 以包括指示欲將第一資料集（資料1)儲存於第一資料儲存器 135210.doc 201005734 件42之第一位置52中 '欲將第二資料集（資料2)儲存於第一 資料儲存11件42之第二位置54巾、欲將第三資料集（資料3) 儲存於第-資料儲存器件42之第三位置56中及欲將第關 資料集（資料N)儲存於第—f料儲存器件42之第N個位置58 中之資訊。 此外’因為亦欲將N個資料集 ’ 乃，· I丨Ά 命1干 ❿ 44及第三資料儲存器件仏，所以資料儲存器件控制器 34(或資料健存裝置32内之其他適當組件）可以在第二資料 儲存器件44中保留第二邊界資訊區或位置“及在第三資料 儲存器件46中保留第三邊界資訊區或位㈣。與儲存於第 一資料儲存器件42中之邊界資訊類似，儲存於第二邊界資 =61:之邊界資訊包括指示欲將第-資料集(資料υ儲存 於第-資料儲存器件44之第一位置62中、欲將第二資料集 (資料2)儲存於第二資料儲存器件44之第二位置64中 第三資料集(資料3)儲存於第二資料儲存器 : 466 =欲將第關資料集(資料_存於第二資料儲:器ί 4之Ν個位置68中之資訊。同樣地，儲存 訊區71中之邊界眘邙白紅4t_ 硬界資 逯界資訊包括、不欲將第一資 於第三資㈣存器件46之第—位置72中、欲將第二資)= 子 (資=於第三資料儲存器件“之第二位、欲： 第^料集（資料3)儲存於第三資料儲存器件Μ之第^ 76中及欲將第_資料集（資料_存於第三資料儲;器^ 46之第N個位置78中之資訊。 貝付儲存|§件 方法_包括將一或多個資料集寫入至資料错存器件之 I352I0.doc •15- 201005734 ㈣μ〇 —旦已將邊界資訊 制·傲左§g/j· °寫入至（例如）在一戋客袖夕咨 科储存器件之内或外部之適 $多個之資 界資訊所交附之-或多個資料集寫入==藉由邊 如，在第-資料儲存器件42内，==器件。例 入(例如)至第一邊界資 -將適虽邊界資訊寫 寫入至第-位置52:=便將第-資料集_) …將第三資料（mi(㈣2)寫人至第二位置 料集（資料N)寫入至㈣個位置巧。—位置56及將第N個資The storage system carefully stores the system from the mirroring system or other suitable redundant data. Including a data_set 32, the capital = 2 includes a data storage device controller 34. The data storage device controller 34 may include a: a boundary information location of the stored information ^ In addition, the boundary information bits within the generation (4) In addition, the data storage, 円) boundary information location 38. The data storage device 32 can be lightly coupled to a plurality of data stores # via the data storage device controller 34, such as a first data storage device 42, a second resource storage unit 44, and a second data storage device. Each of the data storage devices can be a data storage disk or a disk drive, or any other suitable data storage device. Data storage device 32 is configured to interface to a host system (not shown). The host system is also typically coupled to an application (not shown) that generates data to be stored in data storage system 30. The billet storage device 32 and/or the data storage device controller 34 may comprise, in part or in whole, any suitable structure or configuration (e.g., one or more 135210.doc -12-201005734 integrated circuits). In addition, it should be understood that data storage device 32 includes other components, hardware and software (not shown) for use in data storage device 32 and/or data storage device controller 34 for other features and functions not specifically described herein. . All relevant portions of data storage device 32 and/or data storage device control (4) may be partially or fully configured in the form of hardware circuitry and/or other hardware components within a large device or group of components. Alternatively, the data storage device 32 and/or the data storage device controller % portion may be configured in part or in full soft form (eg, as j = and/or one or more logical or computer code sets). In such a configuration, logic or processing instructions are typically stored in memory elements. The memory component is typically transferred to a processor or controller, such as a data storage device controller 34. The controller accesses the necessary instructions from the memory element and executes the instructions or transmits the instructions to the appropriate locations within the data storage device 32. Referring now to Figure 3 (continue to Figure 2), there is shown a block diagram of a method 8 of a data storage system for writing to 圊2 in accordance with the present invention. Method 8 will be described along with the operation of data storage system 30. As part of the redundant data storage system, the data storage device 32 stores the data thus received to each of the data storage devices 42, 44, 46 via the data storage device controller 34. Therefore, if one of the data storage devices becomes unavailable afterwards (for example) due to disk drive failure, data corruption, or other unavailability conditions, it can be accessed from one of the data storage devices that function properly from other data storage devices. data. The method 80 includes the step 82 of storing boundary information, for example, to one or more of the data storage devices and/or to the data storage device 32. Some or all of the 135210.doc •13-201005734', the embodiment of the data storage device, the portion of the data storage device 32, and/or any other suitable location are reserved for boundary information. Storage. For example, a boundary information location or zone above or within two or all of the storage of the boundary information storage device. Alternatively, the corrugation may retain a boundary information location or zone within the data storage device 32 or alternatively to an external location of the data storage device 32 and one or both of the data storage devices 42, 44, 46. As discussed above, the boundary information may refer to the area where the specific data set is not to be written to the data storage device, the start position and the end position of the area to which the data storage device belongs, and the material storage device. The starting position and length of the area to which the data is to be written and/or other suitable information about the data set to be written to the data storage device 42, 44, 46. For example, if N data sets are received by data storage device 32 for storage to each of data storage devices 42, 44, 46, then data storage device controller 34 (or, for example, data storage device 3 2) The other appropriate component) maintains a boundary information area or location for storage of boundary information associated with the storage of N data sets to individual data storage devices and/or to all data storage devices. For example, the data storage device controller 34 maintains a first boundary information area 51 in the first data storage device 42 for storage of boundary information associated with the N data sets stored to the first data storage device 42. Alternatively, the data storage device controller 34 retains the boundary information area 36 therein and/or retains the boundary information area 38 in the data storage device 32 for boundary information associated with the n data sets stored to the first data storage device 42. Store. Such boundary information may include indicating that the first data set (data 1) is to be stored in the first location 52 of the first data storage 135210.doc 201005734, 42. 'To store the second data set (data 2) A data storage 11 piece 42 of the second position 54 towel, the third data set (data 3) is to be stored in the third position 56 of the first data storage device 42 and the first information set (data N) is to be stored in The information in the Nth position 58 of the first material storage device 42. In addition, because the N data sets are also required, the data storage device controller 34 (or other suitable components in the data storage device 32) The second boundary information area or location may be retained in the second data storage device 44 and the third boundary information area or bit (4) may be retained in the third data storage device 46. The boundary information stored in the first data storage device 42 Similarly, the boundary information stored in the second boundary asset = 61: includes indicating that the data set is to be stored in the first location 62 of the first data storage device 44, and the second data set (data 2) is to be The third data set (data 3) stored in the second location 64 of the second data storage device 44 is stored in the second data storage: 466 = the first data set is to be stored in the second data storage: 4, the information in a position 68. Similarly, the boundary in the storage area 71 is cautiously white and red 4t_ hard information is included, and the first is not required to be the third (four) storage device 46 - Position 72, want to be the second capital) = child (in the third data storage device) The second place, the desire: The first collection (data 3) is stored in the third data storage device 第 in the 76 and the _ data set (data _ stored in the third data storage; device ^ 46) Information in N locations 78. Pays for storage | § Method _ includes writing one or more data sets to the data loss device I352I0.doc •15- 201005734 (4) μ〇—has border information system The left §g/j·° is written to, for example, the attached information of one or more of the information within or outside the hacker's storage device - or multiple data sets written == By way of example, in the first data storage device 42, == device. For example, to the first boundary resource - write the boundary information to the first position 52: = the first data set _) ... write the third data (mi ((4) 2) to the second location (data N) to (four) position. - Location 56 and the Nth capital
同樣地，在第二資料儲存器件4艸， 資訊寫入(例如)至第二邊界資訊位置 (資料1)寫入至第一位置62，㈣4更將$資科集 當1番“ 置62 #第二資料集(資料2)寫入至 一 ’將第三資料集(資料3)寫入至第三位置66及將 第N個資料集(資料N)寫入至第,個位置 存器件46中，一曰p此、^ 一已將適备邊界資訊寫入（例如）至第三邊 界貝訊位置71’便將第一資料集（資料υ寫入至第-位置Similarly, in the second data storage device 4, the information is written, for example, to the second boundary information position (data 1), and is written to the first location 62, and (4) 4 is further divided into $. The second data set (data 2) is written to a 'write the third data set (data 3) to the third location 66 and write the Nth data set (data N) to the first location memory device 46. In the middle, the first data set (data υ is written to the first position) is written to the third boundary position 71'.
72將第一資料集(資料2)寫入至第二位置μ，將第三資料 集(資料3)寫入至第三位置76及將第Ν個資料集(資料ν)寫 入至第Ν個位置78。 方法80亦包括移除或刪除已儲存邊界資訊之步驟86。一 旦已完成將-或多個資料集寫人至所有資料儲存器件之步 驟84,便將與正在寫入至資料儲存器件之該資料相關之邊 界-貝訊從適當邊界資訊位置或區移除或刪除。此外，若將 與Ν個資料集儲存至資料儲存器件42、44、牝之一或多個 相關之邊界資訊寫入至資料儲存器件控制器34内之邊界資 135210.doc -16 - 201005734 訊位置36及/或資料儲存裝置32内之邊界資訊位置38，則 資料集寫入至適當資料儲存器件完成時可以自此移除此類 邊界資訊。 方法80亦可包括偵測與回復步驟92。若資料寫入步驟84 期間針對資料儲存系統30發生一中斷（一般顯示為88)，例 如寫入操作中斷，則方法80可以偵測寫入中斷並採取必需 • 的步驟以從寫入中斷回復。依據本發明之具體實施例，從 寫入操作中斷回復使用邊界資訊，因此（例如）藉由比習知 • 技術更快且高效地修復寫入操作中斷之效應而與習知技術 相比改善回復程序。 現在參考圖4(繼續參考圖3)，顯示依據本發明之具體實 施例的圖2之偵測與回復步驟92之方塊圖。偵測與回復步 驟92包括偵測寫入中斷之步驟94。偵測步驟94可以採用任 何適合的方式（例如習知資料儲存方法及系統中實行寫入 操作中斷之偵測所採用之方式）偵測寫入操作中斷。 _ 一已偵'則到寫入操作中斷，價測與回復步驟92便可使 用邊界資訊來支援從中斷寫入操作之相對高效回復。例 如在具體實施例中，回復涉及將臨界資料區複製至已 纟及資料儲存器件之步驟96。如上文中所論述臨界區係 資料储存器件之寫入中斷期間正在寫入所至之區。邊界資 訊定義或以其他方式識別各適當資料儲存器件上為寫入操 乍中斷之接收方之臨界區。依據本發明之具體實施例，僅 ^區得以考量回復與校正。意即，資料儲存器件之在臨 界區外部或與臨界區不同的區域不必加以校正(即，複製 135210.doc 17 201005734 或重寫）’且因此在所有資料儲存器件上保持相同。 旦已使用邊界資訊識別臨界區，複製步驟96便將來自 來源資料儲存器件（即，寫入操作中斷未危及其臨界區之 資料儲存器件）之臨界區資料複製至目標資料儲存器件 (即，寫入操作中斷危及或可能已危及其臨界區之資料儲 存器件）之每一者之臨界區。與校正或重寫已危及（目標）器 件之所有資料區之習知技術相反，藉由僅校正或重寫各目 標器件之臨界區，依據本發明之具體實施例之回復程序比 %知校正技術咼效得多，例如，較少耗時且程序密集。 依據本發明之另一具體實施例，偵測與回復步驟92可以 實行一替代回復程序，其涉及搜尋且決定所有資料儲存器 件之臨界區間之差異（即，來源資料儲存器件（等）之臨界區 與目標資料儲存器件之臨界區間之差異）之步驟98。再 者’依據本發明之具體實施例，邊界資訊係用於臨界區之 識別。僅僅資料儲存器件之臨界區間之差異之此識別與決 定所有資料儲存器件之所有資料儲存區間之差異的某些習 知回復技術相比較。依據僅集中於資料儲存器件之臨界區 上的本發明之具體實施例，一旦已決定此等臨界區間之差 異，偵測與回復步驟92便實行僅將來源資料儲存器件之臨 界區與目標資料儲存器件之臨界區間不同的資料從來源資 料儲存器件複製至目標資料儲存器件之步驟〗〇2。 回復完成時’偵測與回復步驟92實行返回至將一或多個 資料集寫入至資料儲存器件之步驟84之步驟104。方法80 接著如上文中所論述繼續。 135210.doc • 18 · 201005734 熟習此項技術者應明白可對本文所說明的本發明之具體 實施例進行許多改變與替代而不背離由所附申請專利範圍 所定義的本發明之精神與料及其等效者之全部 【圖式簡單說明】 圖1係一包括複數個資料儲存器件之習知冗餘資料儲存 系統之方塊圖’該等資料儲存器件耦合至-包括適於使用 該資料儲存器件之應用程式的主機系統；72, the first data set (data 2) is written to the second position μ, the third data set (data 3) is written to the third position 76, and the third data set (data ν) is written to the third Location 78. The method 80 also includes the step 86 of removing or deleting the stored boundary information. Once the step 84 of writing the data collection to the data storage device has been completed, the boundary associated with the data being written to the data storage device is removed from the appropriate boundary information location or region or delete. In addition, if the boundary information associated with one or more of the data storage devices 42, 44, 牝 is written to the boundary of the data storage device controller 34, the location is 135210.doc -16 - 201005734 36 and/or the boundary information location 38 in the data storage device 32, such boundary information can be removed from the data set when the data set is written to the appropriate data storage device. Method 80 can also include a detect and reply step 92. If an interrupt occurs during data write to step 84 (generally shown as 88), for example, if the write operation is interrupted, then method 80 can detect the write interrupt and take the necessary steps to reply from the write interrupt. In accordance with a particular embodiment of the present invention, the use of boundary information is replied from a write operation interrupt, thereby improving the reply procedure as compared to conventional techniques, for example, by repairing the effects of write operation interruptions faster and more efficiently than conventional techniques. . Referring now to Figure 4 (continuing with reference to Figure 3), a block diagram of the detection and recovery step 92 of Figure 2 in accordance with a particular embodiment of the present invention is shown. The detect and reply step 92 includes a step 94 of detecting a write interrupt. The detecting step 94 can detect a write operation interruption in any suitable manner, such as the conventional data storage method and the manner in which the detection of the write operation interruption is performed in the system. _ Once detected, the write operation is interrupted, and the price measurement and reply step 92 can use the boundary information to support a relatively efficient reply from the interrupt write operation. For example, in a particular embodiment, the reply involves the step 96 of copying the critical data area to the data storage device. The region to which the critical section data storage device is written is interrupted during the write interrupt period as discussed above. The boundary information defines or otherwise identifies the critical section of the recipient of the appropriate data storage device that is the interrupt to the write operation. In accordance with a particular embodiment of the present invention, only the ^ zone is considered for recovery and correction. That is, the area of the data storage device that is external to or different from the critical area does not have to be corrected (i.e., duplicated 135210.doc 17 201005734 or rewritten)' and thus remains the same on all data storage devices. Once the boundary information has been identified using the boundary information, the copying step 96 copies the critical section data from the source data storage device (ie, the data storage device whose write operation is not critical to its critical region) to the target data storage device (ie, write The critical section of each of the data storage devices that interrupt or endanger the critical section. In contrast to conventional techniques for correcting or rewriting all data areas of a compromised (target) device, by simply correcting or rewriting the critical region of each target device, the reply procedure is more specific than the % known correction technique in accordance with an embodiment of the present invention. It is much more effective, for example, less time consuming and procedural intensive. In accordance with another embodiment of the present invention, the detect and reply step 92 can implement an alternate reply procedure that involves searching for and determining the difference in the critical interval of all data storage devices (ie, the critical region of the source data storage device (etc.) Step 98 of the difference from the critical interval of the target data storage device. Further, in accordance with a specific embodiment of the present invention, boundary information is used for identification of critical sections. This identification of only the difference in the critical interval of the data storage device is compared to some conventional recovery techniques that determine the difference in all data storage intervals for all data storage devices. According to a particular embodiment of the invention that is only concentrated on the critical section of the data storage device, the detection and recovery step 92 performs the storage of only the critical region and target data of the source data storage device once the difference in critical intervals has been determined. The steps of copying the different critical sections of the device from the source data storage device to the target data storage device are described in 〇2. Upon completion of the reply, the Detect and Respond step 92 performs a step 104 of returning to step 84 of writing one or more data sets to the data storage device. The method 80 then continues as discussed above. 135210.doc • 18 · 201005734 It will be apparent to those skilled in the art that the present invention may be modified and substituted without departing from the spirit and scope of the invention as defined by the appended claims. BRIEF DESCRIPTION OF THE DRAWINGS [Brief Description of the Drawings] FIG. 1 is a block diagram of a conventional redundant data storage system including a plurality of data storage devices. The data storage devices are coupled to-including a data storage device suitable for use. The host system of the application;
圖2係依據本發明之具體實施例之_資料儲存 塊圖； 乃 圖3係依據本發明之具艘實施例用於寫入至圖 存系統之-方法之方塊圖；& 〈貧枓儲 圖4係依據本發明之具體實施例的圖2之偵測 之方塊圖。 復步驟 【主要元件符號說明】 10 習知冗餘資料儲存系 12 資料儲存器件控制器 14 第一資料儲存器件 16 第二資料儲存器件 18 第二資料储存器件 22 主機系統 24 應用程式 30 冗餘資料儲存系統 32 資料儲存裝置 34 資料儲存器件控制器 135210.doc _ 19- 201005734 36 38 42 44 46 51 52 54 5 6 582 is a block diagram of a data storage block in accordance with an embodiment of the present invention; FIG. 3 is a block diagram of a method for writing to a memory system in accordance with an embodiment of the present invention; & 4 is a block diagram of the detection of FIG. 2 in accordance with an embodiment of the present invention. Complex Steps [Main Component Symbol Description] 10 Conventional Redundant Data Storage System 12 Data Storage Device Controller 14 First Data Storage Device 16 Second Data Storage Device 18 Second Data Storage Device 22 Host System 24 Application Program 30 Redundant Data Storage System 32 Data Storage Device 34 Data Storage Device Controller 135210.doc _ 19- 201005734 36 38 42 44 46 51 52 54 5 6 58
71 72 74 76 邊界資訊位置/邊界資訊區 邊界資訊位置/邊界資訊區 第一資料儲存器件 第二資料儲存器件 第三資料儲存器件 第一邊界資訊區 第一位置 第二位置 第三位置 第N個位置 61 第二邊界資訊區或位置 62 第一位置 64 第二位置 66 第三位置 68 第N個位置 第三邊界資訊區或位置 第一位置 第二位置 第三位置 78 第N個位置 135210.doc -20-71 72 74 76 Boundary Information Location/Boundary Information Zone Boundary Information Location/Boundary Information Zone First Data Storage Device Second Data Storage Device Third Data Storage Device First Boundary Information Zone First Position Second Position Third Position Nth Position 61 Second boundary information area or position 62 First position 64 Second position 66 Third position 68 Nth position Third boundary information area or position First position Second position Third position 78 Nth position 135210.doc -20-